Transforming E-Commerce Logistics: Automating Article Onboarding with Generative AI

About the Event 📌

Scaling e-commerce operations requires smart automation. At Zalando, manually classifying new partner products into "Commodity Groups" created a massive bottleneck, costing roughly $0.10 per review with an 84-hour average turnaround. ⏳

To solve this, we architected an AI-powered image and text processing pipeline powered by multimodal generative AI (Gemini 2.5 Flash). Today, 95% of articles are reviewed automatically in seconds, with only 5% requiring manual handling, scaling to 400 reviews per minute while cutting costs by 97%! ⚡

Join us for a look at how we safely transitioned from a manual bottleneck to production-scale AI automation without disrupting live operations. 🤖

What We Will Cover 🔍
🧠 Prompt Engineering for Real World Data: How we used "Visual Confirmation" and "Functional Redirects" to eliminate AI hallucinations on article categorization.
🛡️ De-risking AI Deployments: Implementing a parallel "Shadow Mode" to prove accuracy against human benchmarks before giving the AI decision authority.
⚙️ Smart Queue Architectures: Building a two-stage status system to eliminate race conditions between AI processors and human reviewers.
📈 Lessons from Scaling: Practical key takeaways for applying enterprise AI to supply chain and logistics processes.

Who Should Attend? 👥
Tech enthusiasts, software architects, product managers, and leaders looking to move AI projects out of the lab and into high-volume production. 💬

Bring your questions and join the discussion! 🙌